Craniofacial microsomia (CFM) involves the asymmetric underdevelopment of facial skeletal bones and soft tissue and occurs in over 1 in 3500 births. Typically affecting the jaw and ear, CFM impairs basic functions such as hearing, speech, respiration, and/or chewing, in addition to its aesthetic effects. It can be life- threatening due to airway compromise. Animal models of teratogenesis have demonstrated a relationship between the features of CFM and vascular disruption (bleeding, clotting, or disruption of blood vessel formation). The only large epidemiologic study of CFM showed associations with maternal exposures that influence the risk of bleeding or clotting, such as pseudoephedrine use and cigarette smoking during pregnancy. CFM-like features have also been observed together with disrupted vasculogenesis in animals lacking genes related to the activities of retinoic acid (a transcription factor) and endothelin (a vasoconstrictor). Common variation in such genes could alter their activity and thereby influence the risk of CFM. Our specific aims are to address the vascular disruption hypothesis of CFM etiology by assessing its association with variation in candidate genes involved in vasculogenesis and hemostasis. We will implement a novel case-parent triad hybrid approach that incorporates data from control parents. We will extract and amplify DMA from existing buccal specimens banked as part of a multicenter case-control study of CFM (172 case-parent and 382 control-parent sets). We will perform single nucleotide polymorphism (SNP) discovery in genes related to retinoic acid activity by resequencing two sample populations. We will select tagSNPs that comprehensively characterize the variation in each of these and other candidate genes (totaling 19) involved in the activities of retinoic acid, endothelin, and thrombin, a key coagulation factor. We will genotype cases, their parents, and the parents of controls by using a high-throughput genotyping platform. We will estimate the association between CFM and offspring and maternal genotypes by fitting log-linear models and perform gene-wide tests of significance. Combining data from case-parent triads and parental controls allows one to capitalize on the advantages offered by the two different study designs and test assumptions of the log-linear model. Epidemiologic, clinical, and teratology studies have strongly supported the vascular disruption hypothesis for CFM. However, further research is necessary to elucidate the pathogenic processes involved in the vascular etiologies of this condition. The proposed research addresses this gap in knowledge through a state-of-the-art tagSNP approach. This will be the first molecular epidemiologic study of CFM. The results of our proposed research could lead to the development of interventions that reduce CFM incidence. The findings should also improve understanding of craniofacial development.